摘要
研究以某市44所高中9610名学生的三年数学成绩数据为基础,对比简单线性回归模型、学生增值百分位模型和多水平模型三种增值评价方法在学校效能评估中的应用。主要分为两部分:第一部分以中考和高考数学成绩为起终点计算增值;第二部分以学年为单位分阶段计算增值,并取各学年平均值作为学校增值结果。结果表明,采用简单线性回归模型与学生增值百分位模型在学校增值排序上高度一致,相关系数为0.988;采用多水平模型的方法因考虑学校层面因素,排序结果与前两种方法略有差异,但变化幅度较小,相关系数分别为0.955和0.956。总体而言,三种方法在学校效能评估中具有较高一致性,但多水平模型方法在捕捉学校层面影响因素上更具优势,能为教育管理部门选择增值评价方法提供依据,并为构建综合教育评价体系提供参考。
This study is based on three years of mathematics achievement data from 9,610 students across 44 high schools in a certain city,comparing the application of three value-added assessment methods—simple linear regression model,student growth percentile model,and multilevel model—in evaluating school effectiveness.The research is divided into two parts:the first part calculates value-added scores using initial(senior high school entrance examination)and final(college entrance examination)mathematics scores;the second part calculates annual value-added scores and takes the average across years as the final school value-added result.The results show a high degree of consistency between the simple linear regression model and the student growth percentile model in school ranking,with a correlation coefficient of 0.988.Although the multilevel model,which accounts for school-level factors,yields slightly different rankings compared to the other two models,the variation is minor,with correlation coefficients of 0.955 and 0.956,respectively.Overall,the three methods demonstrate high consistency in evaluating school effectiveness,but the multilevel model offers greater advantages in capturing the influence of school-level factors.This study provides a reference for educational administrators in selecting appro priate value-added assessment methods and contributes to the construction of a comprehensive educational evaluation system.
作者
游晓锋
杨建芹
YOU Xiaofeng;YANG Jianqin(School of Mathematics and Information Science,Nanchang Normal University,Nanchang 330032,China)
出处
《南昌师范学院学报》
2025年第2期91-96,共6页
Journal of Nanchang Normal University
基金
南昌师范学院校重点项目“增值评价模型在教育实践中的选择和应用研究”(21XJZX28)。
关键词
增值评价
简单线性回归模型
学生增值百分位模型
多水平模型
value-added assessment
simple linear regression model
student growth percentile model
multilevel model
